Research

WORKING PAPERS

How important are matching frictions in the labor market? Experimental and non-experimental evidence from one Indian firm  (with Abhijit Banerjee)

Coverage: PEDL CEPR interview

Abstract: This paper provides evidence for substantial matching frictions in the labor market in India. In particular, placement officers in vocational training institutes have very little information about the job preferences of candidates they are trying to place in jobs. In the first part of this study, we adopt a number of methods to elicit genuine preferences of candidates over different types of jobs and show that placement officers have poor knowledge of these preferences. In the second part, we provide placement officers with this information and examine its impact on placement outcomes and employment. We find that placement officers come close to efficiently matching candidates to interviews subject to their information constraints on preferences. Furthermore, this leads substantial improvement in job choices made by candidates and subsequent employment outcomes for up to six months after initial placement.


Marriage markets and the rise of dowry in India (with Jeff Weaver)

Coverage: Live Mint

Abstract: Dowry payments are an important part of household finances in India, typically exceeding a year of household earnings. Yet there is little empirical evidence on determinants of dowry payments, with existing work relying on small and non-representative samples. In the first part of the paper, we leverage data on over 76,000 marriages since 1900 to document a number of new facts about Indian marriage markets. We show that although there has been little change in some practices, such as intercaste marriage, there are large changes in dowry payments. Between 1945-1975, the proportion of marriages with dowry payments increased from 35-40% to nearly 90%. Over this time period, real dowry payments increased alongside dowry prevalence, but decreased after 1975. In the second half of the paper, we empirically test and reject the major existing theories of dowry, such as changes in sex ratios (marriage squeeze hypothesis), emulation of high caste individuals (Sanskritization hypothesis), and cross-caste competition for grooms. Instead, the changes in dowry prevalence and size are explained by changes in the earnings distribution of grooms within a search model of marriage markets. This has important implications for designing of anti-dowry policy, a major focus of the Indian government.


The effect of storage capacity on price dynamics and production: Evidence from India (with Sabyasachi Das and Shivakumar Venkatraman)

[Draft coming soon]

Abstract: What is the effect of expanding local storage capacity on seasonal variation in market prices of commodities and how does it affect farmers’ production choice? We answer these questions by examining perishable commodities (especially, potato) in the context of India (specially the state of Uttar Pradesh (UP)). We exploit a large cold storage subsidy program of the Indian government that enabled construction of 14.5 million metric tonne of cold storage capacity in a span of about two decades. We build a parsimonious model with risk-neutral farmers and costly storage to show that storage capacity expansion would (i) (non-linearly) reduce seasonal price dispersion, and (ii) (linearly) increase production even in absence of any (aggregate or idiosyncratic) uncertainty or farmer heterogeneity. The empirical results are consistent with the model and are economically significant. The estimates suggest that in an average district in UP, cold storage construction during 2002-2015 contributed to about one-third in the overall reduction in seasonal price dispersion and one-half in the overall increase in the production of potato. We will address the endogeneity concerns in future drafts of the paper by using the discontinuities in the rules for allocating subsidies.


WORK IN PROGRESS

Why do firms depend on referrals? Experimental evidence from India

Misallocation, barriers to entry and multiproduct firms: Evidence from India